keywords:
Bookmark and Share



Front Back
study that looks for prior exposure, risk factor in group w/ dx
case-control study
study that follows group to determine if exposure increases likelihood of of dx
cohort study
study that determines prevalence of dx & risk factors at a given time
cross-sectional study

can establish assoc w/ dx but not causality
prportion of exposed group that has disease v. proportion of unexposed group that does not have disease
odds ratio ad/bc

used w/ case-control study
odds of exposure in pt w/ dx v. odds of exposure in pt w/o dx
exposure odds ratio

used w/ case-control study
proportion of exposed group that develops dx v. proportion of unexposed group that develops dx
relative risk a(c+d)/(a+b)c

used w/ cohort study
clinical trial that assesses safety, toxicity, pharmacokinetics
phase I
clinical trial that assesses efficacy, dosing, adverse effects
phase II
clinical trial that compares new tx to standard of care
phase III
post-marketing surveillance that detects rare, long-term adverse effects
phase IV
TP/(TP+FN)
sensitivity

ability of a test to detect dx when present
TN/(TN+FP)
specificity

ability of a test indicate healthy pt when dx is not present
screening test
sensitivity

used to rule out dx
confirmatory test
specificity

used to rule dx in
TP/(TP+FP)
positive predictive value (PPV)

dependent on dx prevalence
TN/(TN+FN)
negative predicitive value (NPV)
total cases of dx/total population
prevalence
new cases of dx/total population at risk
incidence

ppl w/ dx are not considered at risk
difference in risk that is due to exposure
attributable risk

a/(a+b) - c/(c+d)
reduction in risk assoc w/ tx v. placebo
absolute risk reduction (ARR)

event rate (placebo) - event rate (tx)
number needed to tx (NNT)
1/ARR
number needed to harm (NNH)
1/attributable risk
reduced precision of a test
random error
reduced accuracy of a test
systematic error
nonrandom assignment to study group
selection bias
selecting hospitalized pts as control group
Berkson's bias
information gathered at an inappropriate time
late-look bias
more information gathered from exposed group v. control
detection bias
knowledge of dx alters recall by subjects
recall bias
subjects in different groups are not treated equally
procedure bias
effect of 1 factor distors the effect of another
confounding bias
researcher's belief in tx efficacy changes the outcome of a tx
Pygmalian effect
group being studied changes its behavior once made aware that it is being studied
Hawthorne effect
each subject as as his/her own control
crossover study
mode < median < mean
positive skew

asymmetry w/ tail on right
mean < median < mode
negative skew

asymmetry w/ tail on left
most frequent value
mode

least effected by outliers in a sample
middle value of an ordered number set
median
stating that there is a difference when none exists

"false positive"
Type I error (α)

p = probability of making a type I error
stating that there is no difference when one actually exists

"false negative"
Type II error (β)

inability to detect a difference
likelihood of detecting a difference if it exists
power (1-β)

dependent on sample size
σ/√n

[standard deviation/square root of sample size]
standard error of the mean (SEM)

decreases as sample size increases
mean ± Z(SEM)
confidence interval

95% CI: Z = 1.96
99%CI: Z = 2.58
CI for the mean difference between 2 variables includes 0
conclude no difference
CI for odds ratio or relative risk includes 1
conclude no difference
test that checks difference b/w the means of 2 groups
t-test
test that compares the mean b/w 2 groups, requires population variance
2-sample z-test

(t-test requires sample variance)
test that checks difference b/w the means of 3+ groups
ANOVA
test that checks difference b/w proportions
chi-square (χ2)
coefficient of determination
r2

r = correlation b/w 2 variables, from -1 to 1
x of y cards Next >|